The proposed study is a continuation of a longitudinal investigation of bone mineral loss and nonviolent fractures begun in 1981 on a homogeneous population of 2500 Japanese-American men and women on the island of Oahu, Hawaii. Site-specific bone loss rates are being determined from continuing, serial bone mineral content (BMC) measurements of the lumbar spine, os calcis, distal radius, and proximal radius; serial hip BMC measurements will be initiated. An additional 600 daughters of the cohort in the age range 30-55 years are currently beginning examination. Bone mineral content and site-specific bone loss rates will be correlated with incident fractures, diet, physical activity and dis-use, reproductive variables, drug use (including estrogens and thiazides) and other factors and diseases potentially contributing to bone loss. A major aim of the study is to evaluate the contribution of various risk factors, site-specific BMC, and bone loss rates to fracture risk. Fracture risk in general, irrespective of fracture site, will be evaluated; in addition, the specific-risk of wrist and spine fractures will be examined. The relationship of BMC measurements to hip fracture will be evaluated in a separate cohort of acute hip fractures using case control methods. The effects of discontinuing estrogen and thiazide use will be explored with longitudinal bone loss rates. The mechanism of chronic thiazide effect upon BMC will also be explored in a subset of 50 subjects. Serum will be collected annually on all subjects for future analyses of potential indicators of bone loss. Temporal patterns of the onset and rate of bone loss in the pre- and perimenopausal period will be evaluated. Finally, BMC and bone loss in this cohort will be compared to a native Japanese cohort, using comparable methods, with emphasis upon demographic and dietary differences.